Intraoperative Analgesia Guided by the Nociception Level (NOL) in Laparoscopic Colorectal Surgery: A Preliminary Outcome Study

Introduction: Most of postoperative patients report moderate to severe pain, possibly related to opioid underdosing or overdosing during surgery. Nociception Level (NOL) Index has been proposed for the evaluation of the nociceptionantinociception balance in the perioperative period and by that, may lead to a more appropriate analgesic regimen. NOL-index is scaled from 0 to 100; with previous studies suggesting that values >25 can indicate inadequate analgesia. We designed a NOL-guided and Observational study, 2-cohort (retrospective and prospective) trial to test the hypothesis that protocol-driven intraoperative analgesia guided by NOL during laparoscopic colorectal surgery would improve and reduce titration of intraoperative analgesics. Method: This monocentric, observational, and guided study aimed to compare perioperative data during laparoscopic colorectal surgery with or without the use of NOL monitor (NOL-guided vs. Control group). Intraoperative analgesia was provided by injection of fentanyl, which was performed according to the clinician’s assessment in the Control group or to the NOL monitor in the NOL-guided group. The primary outcome was a reduction in fentanyl consumption during surgery in the NOL-guided group. Results: Between 2017 and 2019, 70 patients (i.e., 36 in the Control group and 34 patients in the NOL-guided group) were included. The mean consumption fentanyl was significantly lower in the NOL-guided group (450mg) versus 600mg in the Control group (P = 0.005). Median postoperative pain score in T0, when the patient woke up in the operation room, was2 (inter-quartile range 0-5) and 0 (0-4) in Control and NOL-guided group, respectively (p=0.132). Conclusion: The NOL index in laparoscopic colorectal surgery decreased with incremental fentanyl doses during sevoflurane anaesthesia. This reduction in the intraoperative analgesia suggests its interesting potential as a nociception monitor during general anaesthesia (GA). Future studies should employ more robust design, be appropriately powered and seek to follow longer term outcomes.

Currently, there are only few clinical outcome studies that assess whether the NOL index impacts GA care during surgery [19,20], and consequently outcome studies using the NOL index were not included in meta-analysis study.
We therefore designed a 2-cohort trial to assess the ability of the NOL monitoring to modify anesthesia care during elective major abdominal surgery. This current clinical study aimed to determine the effect of intra operative NOL-guided fentanyl administration under GA. The primary hypothesis was that intra operative NOL-guided fentanyl administration would reduce fentanyl requirement. The secondary hypotheses were that the NOL-guided group would show reduced in the VAS pain score and in the total morphine consumption in the recovery room when compared to the standard practice.

Method Ethics
This evaluation study was designed as an observational and guided cohort study. Two cohort were analysed from 2017 to 2019: Between April 2017 and October 2018, the retrospective cohort included 36patients that expressed their consent to share their medical information including the physiological parameters recorded during surgery. In this cohort, intravenous analgesia with opioids was guided using standard of care hemodynamic monitoring.
In 2019, 34 patients recruited after the prospective outcome protocol was approved by the Human Ethics Committee in Spain.
All patients received written information about the protocol, had ample time to decide on their participation, and gave oral and written informed consent before enrolment into the study.
The study performed in Complexo Hospitalário Universitário de Ourense, Spain.

Patients
American Society of Anesthesiologists (ASA) class I-IV patients (aged above 17) of either sex, scheduled forlaparoscopic colorectal surgery under general anesthesia were recruited to participate in the study.
Exclusion criteria included inability to give informed consent, pregnancy/lactation, allergy to the opioids used, mean arterial pressure (MAP) at screen ingor on the day of surgery > 160 mmHg or < 55 mmHg, HR at screening or on the day of surgery > 90beats.min -1 or < 45 beats. min -1 , any CNSrelated disorder and body mass index (BMI) above 40.

Study Design
Standard Clinical Care in Complexo Hospitalário Universitário de Ourense, Spain: Before induction of Nociception is intimately related to control of the autonomic nervous system since nociceptive disorders are a primary source of hemodynamic instability, as well as, postoperative chronic pain syndromes [4].
A sufficient level of antinociception has been reached when applied surgical stimuli, clinical responses as heart rate (HR) and blood pressure (BP) elevations no longer occur [1][2][3]. Nevertheless, according to a recent clinical trial, despite absent or mild clinical responses, nociceptive activation persists during deep general anaesthesia (GA). Therefore, lack of clinical responses is not indicative of the absent nociception specific activation [5].
Undoubtedly, opioids are effective antinociceptive agents and one of the three pillars of balanced anaesthesia is the administration of opioids in the perioperative period [1,6].
Although opioids are the most effective antinociceptive drug, they cannot be easily quantified or monitored during GA. Analgesics are administered somewhat theoretically, in the light of population pharmacokinetic models, individual patient comorbidities, and generally adjusted to variations in HR and BP. Despite all this, inter-subject opioid pharmacological variability is very well established in the literature [7,8]. Furthermore, both HR and BP are influenced by other factors, including notably hypnotic anaesthetic agent blood concentration and haemodynamic status, and have been shown to be poor predictors of analgesic requirements [9].
Additionally, opioids that are the most effective antinociceptive drug, have undesirable side effects, such as respiratory depression, pharyngeal muscle weakness, postoperative nausea and vomiting, urinary retention, constipation, ileum, pruritus, tolerance and hyperalgesia that may progress into chronic pain syndrome [6,10,11]. Nausea and vomiting are particularly responsible for delayed patients' recovery, prolonged patient stay in the recovery area and, therefore, delayed hospital discharge. It is also known that opioids disorganize sleep pattern and may lead to postoperative delirium [12].
Furthermore, patients receiving opioids as part of GA and leaving the hospital with opioid prescriptions, appear to have an increased risk of opioid dependence [13]. Therefore, it is debatable whether perioperative opioid administration is appropriate or necessary in current clinical practice [14].
Recently, nociception monitors have been introduced to track nociception during anesthesia and guide administration of analgesics, usually opioids. A reliable nociception monitors may help optimize analgesic management.
The Nociception Level (NOL) monitor (Medasense Biometrics Ltd, Ramat Gan, Israel) is able to reliably detect and quantify mild to intense noxious stimulation during anaesthesia and surgery and outperforms haemodynamic indices (BP, HR) in the ability to distinguish between noxious and non-noxious stimuli [15][16][17]. The NOL monitor usesan algorithm based on advanced statistical and machine learning technologies. The algorithm combines multiple autonomic signals (finger photo-plethysmograph amplitude, skin anesthesia, dexamethasone 4mg and ondansetron 4mgwere given. Patients received an IV-line, BP cuff, Train-of-Four (TOF) for neuromuscular monitoring, and bispectral index (BIS) monitoring (Philips, Eindhoven, the Netherlands) for measurement of depth of anesthesia.
Anaesthesia was induced with 150 mcg fentanyl followed by 2 mg/kg propofol. After consciousness was lostas detected by BIS values<60, absence of eyelash reflex, andno response to verbal stimulation, a neuromuscular blocking agent, rocuronium 0.6 mg/kg, was administered. After fullrelaxation, the trachea was intubated. If HR was less than 60 bpm, atropine 0.01 mg /kg was also administered.
After orotracheal intubation (IOT), patients receive 0.5-3% sevoflurane to maintain a BIS between 40-60 and bolus rocuronium 10% of the initial dose to maintain a TOF of 1-2 train counts of 4.

Maintenance in the Retrospective Cohort -The Control
Group: In the control group (retrospective cohort), fentanyl dosing was dependent solely on hemodynamics (MAP, HR) as NOL-index values were not available. If systolic BP and / or HR increased in 20% from the baseline,150 mcg fentanyl was administered. Vasoactive or vasodilation medication as per the anesthesiologist discretion.

Maintenance in the Prospective Cohort Group -NOL-Guided Group:
In the test group, fentanyl dosing was dependent on the NOL-index. In cases where the NOL index was >25 for at least60sec, 150mcg fentanyl was administered. After fentanyl was given, 5 to 10 min were allowed before the next evaluation took place. In cases where NOL-index decreased below 25, no fentanyl was given.
If NOL-index was below 25 and the MAP was below 55 mmHg, vasoactive medication (ephedrine, phenylephrine, norepinephrine), crystalloids, or both could have been given.

Treatment in the Post-Anesthesia Care Unit (PACU):
Patient`s monitoring and medical treatment were followed by the internal standard practice and were identical in both cohorts. After admission to the PACU, standard monitors were applied measuring heart rate, non-invasive blood pressure and oxygen saturation. Patients in the NOL-guided group connected to the NOL device for monitoring pain during staying at the PACU.
Immediate postoperative pain intensity was assessed in T0 i.e., as soon as the patient awoke from surgery, and at 1 hour interval for the 3 hours of PACU stay by using a 0-10 VAS score (0=no pain and 10=worst pain imaginable), with VAS < 3 corresponding to no or mild pain and VAS >3 corresponding to moderate-to-severe pain. In this study, patients who experienced VAS more than 5, received 2mg IV morphine titration and repeated every 10 min. until pain returned to an VAS of < 5.
The same discharge criteria were applied to both groups: controlled pain, hemodynamic stability and a modified Aldrete test score. The Aldrete score is a composite index with 0-2 points for limb movement, respiration, circulation, consciousness, and oxygen saturation, and reflects the level of recovery from anesthesia. An Aldrete score of 10 indicates full recovery and PACU discharge readiness, although pain and other issues (nausea/vomiting, surgical complications, logistics) will delay discharge from the PACU.

Statistical analyses
Since at the time of writing the protocol, no studies have yet been submitted to assess NOL-guided analgesia, we remained unspecified on effect sizes and SDs. Using our vast knowledge in management of anesthesia, we know that in colorectal surgery, the patients receive about 750mcg of fentanyl IV. For an expected decrease of 60% (300 mcg) in intra operative fentanyl consumption in the NOL-guided group, with a type I error α=0.05 (two-tailed), a 90% power, the total sample size needed was 65 (30percohort). To account for anapproximative5% rate of loss to follow-up or missing data due to technical problems, a total of 70 subjects were recruited.
Numerical variables with normal distribution are presented as mean ± standard deviation (SD), and those without normal distribution are given as median (25 th -75 th ). Categorical variables are expressed in terms of frequency (percentages). Differences between the groups were analysed using the independent samples t-test for numerical variables with normal distribution, and the Mann-Whitney U test was used for those without normal distribution. The relationships between the categorical variables were evaluated using the Chi square test. A probability value of p < 0.05 was considered statistically significant.

Results
We conducted this study with 70 patients (36 in the retrospective cohort and 36 in the prospective cohort), aged above 17 yrs. old between April 2017 and April 2019. Recruitment ended when the number of included patients reached the calculated required sample size. Data from all 70 patients were analyzed. No patients were excluded from the study. There were no significant differences between the patients in terms of demographic data and ASA classifications (Table 1).
When the patients were compared in terms of duration of anesthesia and the surgery performed, there was no difference between the two cohorts (data not shown). Surgical procedure and anesthesia times were similar between treatment groups (data not shown). All study patients completed the trial without any harms or adverse events.

Intraoperative Fentanyl Consumption
The intraoperative fentanyl administration was less in the NOL-guided group compared with the Control group by 25%: mean fentanyl in the NOL-guided 450mcg versus 600mcg in the Control group (median difference 150mcg with 95% CI 150 to 300mcg, Mann-Whitney P=0.005). (Table 2) sum the total fentanyl consumption in the 2-cohort study (Figure 1).

Pain in the PACU
The individual pain scores (VAS) in T0, i.e., immediately after the patient awoke in the OR, were consistently higher in patients who received standard care compared with those who received fentanyl dosing based on the NOL-index but the median and the mean total VAS score found to be insignificant. The median pain score in T0 was0 (inter-quartile range 0-4) and 2 (inter-quartile range 0-5) in NOL-guided and Control groups, respectively (P=0.132, Table 3).

Morphine Consumption in the PACU
Cumulative morphine in the PACU did not differ between groups (NOL-guided group 5mg, Control group4mg, mean difference 1mg, p=0.953).In both groups the Aldrete score reached values > 8 in the majority of patients in the PACU. The mean time spent in the PACU was 254 min. in the Control group and 247 min. in the NOL-guided group (p=0.737). No awareness was noted in the PACU or later on the ward (Table 4).

Discussion
Most of the patients in the PACU are characterized with a number of physiological disturbances caused by emergence from anesthesia and surgery, which affect multiple organs and systems. Postoperative pain and following agitation usually add risks and deterioration to complications with their interactions, aggravating outcomes. According to the guidelines by the American Society of Anesthesiologists, routine assessment and monitoring of pain detects complications and reduces adverse outcomes, which should be performed during emergence and recovery. However, unconsciousness and/or inabilities of clear verbal expression bring more difficulties in assessment and treatment for pain in the PACU.
Intraoperative high-dose opioid consumption is associated with postoperative hyper algesia and can be associated with delays in recovery, postoperative nausea, and vomiting, which may extend time to discharge from the recovery unit.
Hence, given all of the above, it is important to prevent high pain levels during and after surgery. As both opioid under dosing and overdosing during anaesthesia are associated with high postoperative pain scores, we designed a study in which we examined whether opioid dosing based on the individual nociceptive state of the patient during surgery, as measured by the NOL index, would improve postoperative pain scores. Our study shows that NOL monitoring statistically significantly lowers opioid consumption in patients undergoing colorectal surgery, under general anesthesia.
Currently, several nociception monitors are available for detection of nociceptive events during surgery. The algorithms used to derive the level of nociception differ among monitors and most monitors measure single physiological variables such as HR variability or combine two variables such as BP and HR. In contrast to the other nociception monitors, the multiparameter NOL-index that we used has an algorithm based on machine learning technology. Earlier studies indicate that the NOL-index outperforms other indices of nociception including BP, HR, the pulse-plethysmograph amplitude and the surgical pleth-index (GE, Healthcare, Helsinki, Finland; an index that combines pulsebeat interval and pulse wave amplitude), in discriminating noxious from non-noxious stimuli [15][16][17]. Additionally, Meijer F et al. recently showed that NOL-guided remifentanil dosing results in improved haemodynamic stability compared with the Control group [19].
In the NOL-guided group we used the NOL-index absolute cut-offs value of 25. We based this threshold on previous outcome and validation studies with the NOL index [15,19].
The management of intraoperative nociception/ antinociception of the patients in the Control group, and hence, opioid consumption, were attempted to be determined using hemodynamic data in our study.
Although hemodynamic data provide us some ideas about nociception/antinociception, it is not possible to state that they can precisely reflect nociception/antinociception. It is well known that hemodynamic are associated with several factors such as fluid status, patient age, and depth and type of anesthesia. We aimed in our study to decrease the total consumption of fentanyl by 25% by adjusting its dose based on changes observed in the NOL-index measurements in the NOLguided group. The reason can be that the anesthesiologist in the NOL-guided group, is more attentive and keen on keeping the NOL-index between the nociception -antinociception balance of 25 and according to the protocol.
These results indicate that NOL monitoring during anesthesia has certain advantages in terms of opioid consumption. Further studies with larger sample sizes are needed to address whether nociception monitoring results in improved hemodynamic stability. During the patients stay in our clinic none of them developed any major complications. Still, we did not follow our patients beyond their hospital stay and hence remain uninformed on possible long-term complications.
Postoperative pain is the most common concern of patients undergoing surgery [21]. Up to 40% will experience severe pain in the recovery period, with a subsequent increase in the risk of developing chronic pain [22,23]. Our study found that in NOL-guided fentanyl dosing group rather than dosing based on hemodynamic indices (BP and HR) in the Control group, results in a 2-points reduction in VAS pain score in the OR, when the patient immediately woke up after the surgery. This current observation of reduced pain scores in patients receiving NOL-guided fentanyl dosing indicate the added value of a multi parameter monitor driven by an artificial intelligence algorithm based on advanced statistical and machine learning technologies. This reduction was not statistically significant due to the limited number of patients in the study but according to many studies on the relevance of postoperative pain score, this change seems to be clinically significant and therefore its importance is much greater. We argue with the possibility that the low pain experience is due to the remnants of opioids given during surgery because the cumulative amount of opioids in the NOL-guided group was significantly lower than in the Control group.  Open Access | Page 280 | Cumulative morphine in the PACU did not differ between cohort, albeit the absolute values show 50% less morphine use in the NOL-guided group (NOL-guided group 4mg vs. Control group, 2mg, mean difference 2mg; P=0.001).

Limitations
The study has several limitations: [1] In common with all studies on the influence of nociception monitor-guided anesthesia on various anesthesia endpoints, our study was single blinded. This may have influenced the outcome due to an implementation or performance bias.
[2] The sample size was relatively small, but our intention was to detect a relatively large difference in intraoperative fentanyl consumption that are clinically relevant. We observed a difference of 25% points between subjects who had received NOL-guided fentanyl and the control group. These differences are clinically and statistically relevant. [3] We relied on intermittent blood pressure measurements to guide anesthesia in the standard clinical care group. It may well be that a continuous blood pressure signal from an arterial line may have improved our ability to guide anesthesia with more rapid response to nociceptive input [4] This study was powered for single primary objective and some of the secondary objectives evaluated in this study might have shown no difference between groups because of lack of power and too small number of patients. Studies with larger number of patients are necessary to better evaluate these specific objectives.

Conclusions
We studied the influence of nociception-guided anesthesia using the NOL monitor and observed that, compared with standard clinical care, NOL guidance resulted in a statistically significant reduction in the doses of fentanyl administered by 25% in the NOL-guided group compared with the standard of care in laparoscopic colorectal surgery. Furthermore, the patients in this study experience less pain when opioid dosing is guided by the NOL monitor compared with standard of care. These changes resulted in clinically relevant rather than statistically significance.
NOL monitoring had no effect on postoperative opioid requirements.
Future studies in larger more diverse populations, under uncontrolled conditions, should address the long-term outcome benefit of intraoperative analgesia guidance by the NOL index.